Multi-modal Internet of Vehicles congestion propagation optimization method based on niche genetic algorithm

Abstract

The invention discloses a multi-mode Internet of Vehicles congestion propagation optimization method based on a niche genetic algorithm. The method comprises the following steps: acquiring Internet of Vehicles information; the method comprises the following steps: performing mathematical modeling on an Internet of Vehicles congestion propagation influence maximization problem by using Internet of Vehicles information to respectively obtain a congestion propagation dynamic mechanism and an objective function of maximizing the number of Internet of Vehicles congestion road sections; a plurality of heterogeneous congested road section combinations capable of maximizing the number of the congested road sections of the Internet of Vehicles are calculated by adopting a genetic algorithm of an ecological niche technology based on fitness sharing and taking maximization of the number of the congested road sections of the Internet of Vehicles as an optimization target. According to the method, the problem that the existing method is limited to only generate a single road section combination which maximizes the influence of the road network congestion state is solved, and a plurality of heterogeneous road section combinations with excellent performance can be provided based on the genetic algorithm of the niche technology of fitness sharing and the problem of maximizing the influence of a single evolution process on the congestion propagation of the Internet of Vehicles.

Description

Technical Field

[0001] The present invention relates to the technical field of multimodal vehicle network congestion optimization, and in particular to a multimodal vehicle network congestion propagation optimization method based on a niche genetic algorithm. Background Art

[0002] The existing traffic management system faces many challenges. For example, with the annual increase in the number of motor vehicles, the traffic volume has increased, and the optimization of transportation infrastructure is a long process, which has led to the increasingly serious traffic congestion problem in the existing Internet of Vehicles. How to alleviate or avoid congestion in the Internet of Vehicles is an urgent problem to be solved.

[0003] Due to the limited cost of infrastructure, the optimization construction of the existing traffic road network is time-consuming and sequential. Therefore, if we want to improve the existing congestion of the Internet of Vehicles, we can first consider finding several key sections in the existing road network that can maximize the diffusion of road congestion, and optimize these sections to improve traffic. Secondly, decision makers often hope to have multiple such key sections for them to make more scientific road network optimization decisions, which requires solving the multi-modal Internet of Vehicles congestion optimization problem. However, the optimization of the multi-modal Internet of Vehicles congestion problem is a complex problem, with the characteristics of complex road network structure, abstract dynamic mechanism of congestion state diffusion, and multi-peak solution space. Traditional optimization algorithms are difficult to solve it effectively.

[0004] Evolutionary computing combined with microhabitat technology is often used to solve multimodal complex optimization problems and is widely used in various fields. For the problem of maximizing the influence of congestion propagation in multimodal vehicular networks, this method can efficiently find peaks in the multi-peak solution space, thereby providing multiple different but excellent solutions. However, the diffusion mechanism of the congestion state of the road section in the vehicular network is complex, and it is necessary to combine the propagation of complex network states with traffic flow modeling. When applying evolutionary computing to solve the problem of maximizing the influence of congestion propagation in the vehicular network, it is necessary to design a fitness evaluation index that can help the selection mechanism accurately select the optimal solution. At the same time, there is currently no research on maximizing the influence of congestion propagation in the vehicular network from the perspective of multimodal optimization. It is necessary to select appropriate microhabitat technology in the context of this problem to more accurately find multiple significant peaks in the solution space. Therefore, how to apply evolutionary computing combined with microhabitat technology to find multiple heterogeneous solutions to the problem of maximizing the influence of congestion propagation in the vehicular network is a problem that still needs to be solved. Summary of the invention

[0005] In order to overcome the defects and shortcomings of the prior art, the present invention provides a multimodal vehicle network congestion propagation optimization method based on a niche genetic algorithm. The present invention encodes a fixed number of road section combinations as individuals in the genetic algorithm, and uses the road and traffic flow characteristics in the vehicle network to design a fitness function for evaluating the congestion propagation influence of each individual, which can well quantify the propagation influence of the individual; on this basis, a niche technology operator based on fitness sharing is designed, and the operator is embedded in a classic genetic algorithm. After multiple evolutionary iterations of the algorithm, each niche can occupy a different peak in the solution space of the problem of maximizing the congestion propagation influence of the vehicle network, thereby ultimately generating multiple heterogeneous road section combinations that can maximize the congestion propagation influence, so as to solve the problem of solving the road section combination that maximizes the congestion influence of the vehicle network from the perspective of the multimodal optimization problem that is not considered in the prior art, thereby being limited to generating only a single road section combination that maximizes the impact of the road network congestion state.

[0006] The second object of the present invention is to provide a multi-modal vehicle network congestion propagation optimization system based on a niche genetic algorithm;

[0007] A third object of the present invention is to provide a computer readable storage medium;

[0008] A fourth object of the present invention is to provide a computer device.

[0009] In order to achieve the above object, the present invention adopts the following technical solutions:

[0010] The present invention provides a multi-modal vehicle network congestion propagation optimization method based on a niche genetic algorithm, comprising the following steps:

[0011] Obtain Internet of Vehicles information;

[0012] The problem of maximizing the influence of vehicle network congestion propagation is mathematically modeled using vehicle network information, and the congestion propagation dynamic mechanism and the objective function of maximizing the number of congested sections in the vehicle network are obtained respectively;

[0013] A genetic algorithm based on fitness sharing niche technology is adopted, with maximizing the number of congested sections in the vehicle network as the optimization goal, and multiple heterogeneous congested section combinations that can maximize the number of congested sections in the vehicle network are calculated.

[0014] As a preferred technical solution, the Internet of Vehicles information includes road network structure, road capacity and vehicle traffic flow data.

[0015] As a preferred technical solution, the genetic algorithm based on the niche technology of fitness sharing is used to maximize the number of congested sections in the vehicle network, and multiple heterogeneous combinations of congested sections that can maximize the number of congested sections in the vehicle network are calculated. The specific steps include:

[0016] Take all sets containing k vehicle network road section numbers as individuals, and initialize the database, including decision variables and corresponding fitness values, that is, the set containing k vehicle network road section numbers and congestion propagation influence, and set the evolution termination condition;

[0017] Select NP individuals with the best fitness value from the database to form a population, and calculate the congestion propagation influence of each individual in the population as its fitness value based on the congestion propagation influence of the road section;

[0018] Based on the fitness sharing niche method, the population is divided into n sub-populations as different niches according to the similarity of the vehicle network road section combination contained in the individual in the vehicle network topology structure;

[0019] Randomly select two niches from the divided multiple niches and apply the sorted semi-crossover operator;

[0020] A small habitat is randomly selected, and a single-point mutation operator is applied to all individuals in the population with a set probability. If the random number corresponding to an individual is greater than its single-point mutation threshold, a section that does not belong to the individual is randomly selected from the vehicle network section set to replace any section included in the individual.

[0021] Update the fitness values ​​of all individuals in the population, select the individuals with the highest fitness value from each niche to be retained for the next generation, and clone and reproduce a population of size NP;

[0022] The evolution is iterated cyclically until the evolution termination condition is reached, and the individual with the highest fitness in each microhabitat in the population is output as the multimodal optimization solution to the multi-solution optimization problem of the influence of congestion propagation in the Internet of Vehicles.

[0023] As a preferred technical solution, the congestion propagation influence of each individual in the population is calculated based on the congestion propagation influence of the road section as its fitness value. The specific calculation formula is expressed as:

[0024]

[0025] in, It is road section x i In the topology of the Internet of Vehicles, It is road section x i The traffic carrying capacity It is road section x i The number of lanes, It is road section x i , L is the length of the lane occupied by the average motor vehicle for safe driving;

[0026] The individual congestion propagation influence is obtained as:

[0027] As a preferred technical solution, according to the similarity of the vehicle network road section combination contained in the individual in the vehicle network topology structure, the population is divided into n sub-populations as different microhabitats, specifically including:

[0028] Apply community detection algorithm to the Internet of Vehicles and divide it into m Leiden communities;

[0029] For each individual in the population, solve its Internet of Vehicles community belonging vector, the m-dimensional community belonging vector of individual i Among them, the a-th dimension c a represents the total number of road sections that the individual belongs to the connected vehicle community j;

[0030] Based on the individual Internet of Vehicles community belonging vector, the distance metric between individuals is defined, and the Euclidean distance between individual i and individual j

[0031] The IoV belonging vector of each individual is used as a cluster. According to the distance metric between individuals, the clusters with the shortest distance are merged into one cluster until n clusters are formed. Finally, the population is divided into n sub-populations as different microhabitats.

[0032] As a preferred technical solution, two microhabitats are randomly selected from the divided multiple microhabitats and a sorted semi-crossover operator is applied, specifically including:

[0033] Randomly select two individuals i and j from the population, and sort the road sections included in individual i in ascending order and the road sections included in individual j in descending order according to the influence value of road section congestion propagation;

[0034] Exchange the first half of the road section of individual i and individual j to form two new individuals i′, j′;

[0035] At this time, if there are repeated sections in the new individual i′, j′, the repeated sections are removed, and sections are randomly selected from the vehicle network section set to fill in the individual i′, j′ to ensure that the individual i′, j′ still contains k non-repeated sections.

[0036] The present invention also provides a multi-modal vehicle network congestion propagation optimization system based on a niche genetic algorithm, which is used to implement the above-mentioned multi-modal vehicle network congestion propagation optimization method based on a niche genetic algorithm, and the system includes: a vehicle network information acquisition module, a maximization problem model construction module, and a congestion optimization module;

[0037] The Internet of Vehicles information acquisition module is used to acquire Internet of Vehicles information;

[0038] The maximization problem model building module is used to mathematically model the problem of maximizing the influence of vehicle network congestion propagation based on vehicle network information, and obtain the congestion propagation dynamic mechanism and the objective function of maximizing the number of vehicle network congested sections respectively;

[0039] The congestion optimization module is used to adopt a genetic algorithm based on a fitness sharing niche technology, with maximizing the number of congested sections in the vehicle network as the optimization goal, and calculate multiple heterogeneous congested section combinations that can maximize the number of congested sections in the vehicle network.

[0040] The present invention also provides a computer-readable storage medium storing a program, which, when executed by a processor, implements the multimodal vehicle network congestion propagation optimization method based on a niche genetic algorithm as described above.

[0041] The present invention also provides a computer device, comprising a processor and a memory for storing a program executable by the processor, wherein when the processor executes the program stored in the memory, the multimodal vehicle network congestion propagation optimization method based on a niche genetic algorithm as described above is implemented.

[0042] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0043] (1) The present invention uses the acquired Internet of Vehicles data as a mathematical model for establishing the Internet of Vehicles congestion propagation mechanism, quantifies the Internet of Vehicles congestion propagation dynamics mechanism, takes the road section combination as the decision variable, takes maximizing the number of Internet of Vehicles congested sections as the objective function, and adopts the microhabitat technology based on fitness sharing to locate multiple peaks in the solution space of the Internet of Vehicles congestion influence maximization problem, thereby solving the problem that the prior art does not consider the Internet of Vehicles congestion influence maximization problem from the perspective of multimodal optimization.

[0044] (2) The present invention uses the community detection algorithm as an aid, utilizes the community characteristics presented by the IoV road structure, assigns IoV community belonging vectors to all individuals in the genetic algorithm population, and quantifies the differences in IoV road segment combinations at the community granularity, thus solving the problem that the prior art is difficult to reasonably quantify the degree of differences in IoV road segment combinations when applying the microhabitat technology based on fitness sharing. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of a multi-modal vehicle network congestion propagation optimization method based on a niche genetic algorithm of the present invention;

[0046] Figure 2 A schematic diagram of the implementation architecture of the present invention for obtaining the optimal road section combination based on the niche technology and population evolution;

[0047] Figure 3 It is a schematic diagram of the implementation architecture of the microhabitat technology population evolution iteration based on fitness value sharing of the present invention;

[0048] Figure 4 A schematic diagram of the implementation architecture of the microhabitat technology based on fitness value sharing of the present invention;

[0049] Figure 5 A schematic diagram of the semi-crossover operator for sorting of the present invention;

[0050] Figure 6 Schematic diagram of the single-point mutation operator for sorting of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0052] Example 1

[0053] like Figure 1 As shown, this embodiment provides a multi-modal vehicle network congestion propagation optimization method based on a niche genetic algorithm, a genetic algorithm based on a fitness sharing niche technology and a plurality of heterogeneous and excellent road section combinations that can be provided for the vehicle network congestion propagation influence maximization problem in a single evolution process, and a vehicle network community belonging vector is assigned to individuals in the genetic algorithm population by utilizing the community characteristics presented by the vehicle network topology structure, through which the difference between the road section combinations represented by the individuals can be well evaluated, and the road section congestion propagation influence is constructed by utilizing the key factors that cause vehicle network road section congestion, and the algorithm can be efficiently induced to find a higher peak in the solution space of the vehicle network congestion propagation influence maximization problem;

[0054] The method specifically comprises the following steps:

[0055] S1: Obtain Internet of Vehicles information, including road network structure, road capacity and vehicle traffic flow data;

[0056] S2: Mathematically model the problem of maximizing the influence of vehicle network congestion propagation based on vehicle network information, and obtain the congestion propagation dynamic mechanism and the objective function of maximizing the number of congested sections in the vehicle network respectively;

[0057] S3: Figure 2 As shown, a genetic algorithm based on the niche technology of fitness sharing is adopted, with maximizing the number of congested sections in the vehicle network as the optimization goal, and multiple heterogeneous congested section combinations that can maximize the number of congested sections in the vehicle network are calculated.

[0058] like Figure 3 As shown, after the niche technology and population evolution, multiple road section combinations that maximize the congestion propagation of the vehicle network are found. Step S3 specifically includes the following steps:

[0059] S31: taking all sets containing k vehicle network road section numbers as individuals, and initializing the database, including decision variables and corresponding fitness values, that is, the set containing k vehicle network road section numbers and congestion propagation influence, and setting the evolution termination condition;

[0060] In this embodiment, a random sampling method is used to initialize the database to obtain a database with a relatively uniform distribution in the problem solution space. Among them, k represents the number of vehicle network road sections that the individual contains, and each individual represents a potential combination of road sections that can maximize the impact of vehicle network congestion propagation;

[0061] S32: Select NP individuals with the best fitness value from the database DB to form a population in, Represents the decision variable. Based on the congestion propagation influence of the road section, the congestion propagation influence of each individual in the population is calculated, that is, its fitness value. The specific calculation formula is expressed as:

[0062]

[0063] in, It is road section x i In the topology of the Internet of Vehicles, It is road section x i The traffic carrying capacity It is road section x i The number of lanes, It is road section x i , L is the length of the lane occupied by the average motor vehicle for safe driving;

[0064] Then we can get the individual congestion propagation influence

[0065] S33: Applying the microhabitat method based on fitness sharing, the population is divided into n subpopulations, i.e., microhabitats, according to the similarity of the combination of IoV road segments contained by the individuals in the IoV topology structure, so that each microhabitat occupies a peak in the solution space of the IoV congestion propagation influence maximization problem, where the individuals in each microhabitat contain similar IoV road segments at the community granularity of the IoV network topology;

[0066] like Figure 4 As shown, for all individuals in the population, according to the community characteristics presented by the road topology structure of the Internet of Vehicles, a vehicle network community belonging vector is constructed to quantify the composition of the vehicle network road segment combination represented by the individual, and a difference (or similarity) metric of the vehicle network community belonging vector is defined, which can well evaluate the difference between the road segment combinations represented by the individual. Step S33 specifically includes:

[0067] (1) Apply the community detection algorithm to the Internet of Vehicles and divide it into m Leiden communities;

[0068] (2) For each individual in the population, solve its Internet of Vehicles community belonging vector, the m-dimensional community belonging vector of individual i Among them, the a-th dimension c a represents the total number of road sections that the individual belongs to the connected vehicle community j;

[0069] (3) Based on the individual IoV community belonging vector, define the distance metric between individuals, the Euclidean distance between individual i and individual j

[0070] (4) Based on the idea of ​​hierarchical clustering, the IoV belonging vector of each individual is used as a cluster. According to the distance metric between individuals, the clusters with the shortest distance are merged into one cluster until n clusters are formed.

[0071] (5) The population is divided into n sub-populations, which serve as different microhabitats to complete the division of the population's ecological niche.

[0072] S34: in order to accelerate the peak in the niche positioning multi-solution space, taking the niche as a unit, two niches are randomly selected from the divided multiple niches and a sorted semi-crossover operator is applied;

[0073] In this embodiment, step S34 specifically includes:

[0074] (1) Randomly select two individuals i and j from the population. According to the influence value of the congestion propagation of the road segment, the road segments included in individual i are sorted in ascending order, and the road segments included in individual j are sorted in descending order;

[0075] (2) Exchange the first half of the road section of individual i and individual j to form two new individuals i′, j′;

[0076] (3) At this time, if there are repeated road sections in the new individual i′, j′, the repeated road sections are removed, and road sections are randomly selected from the vehicle network road section set to fill in the individual i′, j′ to ensure that the individual i′, j′ still contains k non-repeated road sections;

[0077] Specifically, this embodiment uses the characteristic that each individual is a collection of vehicle network sections, and sorts the sections included in the randomly selected individuals i and j in the population in ascending and descending order respectively according to the influence of the congestion propagation of the sections, and then crosses them;

[0078] For example: In, v i1 The congestion propagation influence of the road section is the greatest, v ikThe influence of road congestion propagation is the smallest, while In, v j1 The congestion propagation influence of the road section is the smallest, v jk The road congestion has the greatest impact on transmission. and In the first half of and And use the road sections in the vehicle network road section collection that do not belong to the new individual to replace the possible repeated road sections caused by the intersection

[0079] like Figure 5 As shown, it is the semi-crossover operator for sorting in this embodiment, wherein the rectangle with the largest area represents the population, and the rectangle with the smallest area represents the microhabitat obtained by the microhabitat technology based on fitness sharing. Microhabitats with gradient colors represent semi-crossover operators that are randomly selected for sorting; the gradient direction represents the sorting method: the gradient from orange to purple indicates that the set of vehicle network sections included in the individual is sorted in ascending order from low to high according to the congestion influence of the section, and the gradient from purple to orange indicates that the set of vehicle network sections included in the individual is sorted in descending order from high to low according to the congestion influence of the section; the new microhabitat obtained by the crossover is marked with a pure color. All individuals in these two microhabitats are semi-crossed, that is, the first half of the combination of the sorted sections are exchanged with each other, thereby forming two new microhabitats. The pure purple rectangle represents a microhabitat containing a combination of sections with higher congestion influence, and the pure orange rectangle represents a microhabitat containing a combination of sections with lower congestion influence. It is worth mentioning that Figure 5 It does not show how to solve the problem of individuals in the new population containing repeated sections that may be caused by the intersection. For this problem, the repeated sections contained in the individual will be replaced by any section in the vehicle network section set that does not belong to the individual.

[0080] S35: To increase population diversity, a small habitat is randomly selected, and a single-point mutation operator is applied to all individuals in the population with a certain probability. If the random number corresponding to an individual is greater than its single-point mutation threshold, a road section that does not belong to the individual is randomly selected from the vehicle network road section set to replace any road section included in the individual;

[0081] In this embodiment, step S35 specifically includes:

[0082] (1) Define the single-point mutation probability threshold p m =0.7, generate random numbers p∈(0,1) for all individuals in the population;

[0083] (2) If the random number p of individual i i >p m , then a road segment that does not belong to individual i is randomly selected from the set of vehicle network road segments and replaced with any road segment in individual i; otherwise, no change is made;

[0084] like Figure 6 As shown, the single-point mutation operator sorted in this embodiment, where the pure orange rectangle represents the initial individual that has not mutated, and the purple-orange gradient rectangle represents the individual that has undergone single-point mutation: all individuals in the population have a certain probability of undergoing single-point mutation. If the randomly generated value exceeds the mutation probability threshold of the individual, a section is randomly selected from the Internet of Vehicles sections included in the individual and replaced by any section in the Internet of Vehicles section set that does not belong to the individual.

[0085] S36: Update the fitness values ​​of all individuals in the population, select the individuals with the highest fitness value from each niche to be retained for the next generation, and clone and reproduce a population of size NP;

[0086] S37: cyclically executing step S33, step S34 and step S35 until the evolution termination condition is reached, that is, the number of evolution iterations reaches the upper limit set by the algorithm;

[0087] S38: Output the individuals with the highest fitness in each microhabitat in the population at this time as the multimodal optimization solution to the multi-solution optimization problem of the influence of congestion propagation in the Internet of Vehicles. On the contrary, if the number of evolutionary iterations has not reached the upper limit set by the algorithm at this time, these individuals will be retained in the next generation of evolution, and they will be used as parents to generate a new population of size NP.

[0088] This embodiment adopts a microhabitat technology based on fitness value sharing. According to the similarity of the topological structure space of the Internet of Vehicles, the population is dynamically divided into n microhabitats. The Internet of Vehicles road segments contained in the individuals in the microhabitats have high similarity in composition, while the Internet of Vehicles road segments contained in the individuals between the microhabitats have low similarity in composition. Therefore, each microhabitat occupies a different peak in the solution space of the problem of maximizing the influence of Internet of Vehicles congestion propagation. Secondly, the sorted semi-crossover operator is applied to the microhabitats as units. Then, the single-point mutation operator is applied to all individuals in the population with a certain probability. Finally, the individual with the highest fitness value is selected from each microhabitat to be retained in the next generation of evolution and cloned into NP individuals as the parent. If the algorithm reaches the termination condition at this time, that is, the number of evolutionary iterations reaches the upper limit set by the algorithm, these individuals are output as solutions to the multi-solution problem of maximizing the influence of congestion propagation in the Internet of Vehicles; conversely, if the number of evolutionary iterations has not reached the upper limit set by the algorithm at this time, these individuals will be retained in the next generation of evolution, and with them as the parent generation, a new population of size NP is generated to continue evolving. This embodiment can be applied to the Internet of Vehicles in different cities. Its goal is to find a plurality of road section combinations with different compositions that can maximize the congestion propagation in the Internet of Vehicles, so as to facilitate the Internet of Vehicles optimization decision makers to select the most appropriate road section combination and carry out optimization at the road engineering level.

[0089] Example 2

[0090] This embodiment provides a multi-modal vehicle network congestion propagation optimization system based on a niche genetic algorithm, which is used to implement the multi-modal vehicle network congestion propagation optimization method based on a niche genetic algorithm in Example 1. The system includes: a vehicle network information acquisition module, a maximization problem model construction module, and a congestion optimization module;

[0091] In this embodiment, the Internet of Vehicles information acquisition module is used to acquire Internet of Vehicles information;

[0092] In this embodiment, the maximization problem model building module is used to mathematically model the problem of maximizing the influence of vehicle network congestion propagation with vehicle network information, and obtain the congestion propagation dynamic mechanism and the objective function of maximizing the number of vehicle network congested sections respectively;

[0093] In this embodiment, the congestion optimization module is used to adopt a genetic algorithm based on fitness sharing niche technology, with maximizing the number of congested sections in the vehicle network as the optimization goal, and calculate multiple heterogeneous congested section combinations that can maximize the number of congested sections in the vehicle network.

[0094] Example 3

[0095] This embodiment provides a storage medium, which may be a ROM, RAM, disk, CD or other storage medium, and the storage medium stores one or more programs. When the program is executed by the processor, the multimodal vehicle network congestion propagation optimization method based on the niche genetic algorithm of embodiment 1 is implemented.

[0096] Example 4

[0097] This embodiment provides a computing device, which can be a desktop computer, a laptop computer, a smart phone, a PDA handheld terminal, a tablet computer or other terminal device with a display function. The computing device includes a processor and a memory, and the memory stores one or more programs. When the processor executes the program stored in the memory, the multimodal vehicle network congestion propagation optimization method based on a niche genetic algorithm of Example 1 is implemented.

[0098] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A multimodal vehicle network congestion propagation optimization method based on niche genetic algorithm, characterized in that: The steps include: Obtain Internet of Vehicles information; The problem of maximizing the influence of vehicle network congestion propagation is mathematically modeled using vehicle network information, and the congestion propagation dynamic mechanism and the objective function of maximizing the number of congested sections in the vehicle network are obtained respectively; A genetic algorithm based on fitness sharing niche technology is adopted, with maximizing the number of congested sections in the vehicle network as the optimization goal, and multiple heterogeneous congested section combinations that can maximize the number of congested sections in the vehicle network are calculated.

2. The multimodal vehicle network congestion propagation optimization method based on niche genetic algorithm according to claim 1 is characterized in that: The Internet of Vehicles information includes road network structure, road capacity and vehicle traffic flow data.

3. The multimodal vehicle network congestion propagation optimization method based on niche genetic algorithm according to claim 1 is characterized in that: The genetic algorithm based on the niche technology of fitness sharing is used to maximize the number of congested sections in the vehicle network, and multiple heterogeneous combinations of congested sections that can maximize the number of congested sections in the vehicle network are calculated. The specific steps include: Take all sets containing k vehicle network road section numbers as individuals, and initialize the database, including decision variables and corresponding fitness values, that is, the set containing k vehicle network road section numbers and congestion propagation influence, and set the evolution termination condition; Select NP individuals with the best fitness value from the database to form a population, and calculate the congestion propagation influence of each individual in the population as its fitness value based on the congestion propagation influence of the road section; Based on the fitness sharing niche method, the population is divided into n sub-populations as different niches according to the similarity of the vehicle network road section combination contained in the individual in the vehicle network topology structure; Randomly select two niches from the divided multiple niches and apply the sorted semi-crossover operator; A small habitat is randomly selected, and a single-point mutation operator is applied to all individuals in the population with a set probability. If the random number corresponding to an individual is greater than its single-point mutation threshold, a section that does not belong to the individual is randomly selected from the vehicle network section set to replace any section included in the individual. Update the fitness values ​​of all individuals in the population, select the individuals with the highest fitness value from each niche to be retained for the next generation, and clone and reproduce a population of size NP; The evolution is iterated cyclically until the evolution termination condition is reached, and the individual with the highest fitness in each microhabitat in the population is output as the multimodal optimization solution to the multi-solution optimization problem of the influence of congestion propagation in the Internet of Vehicles.

4. The multimodal vehicle network congestion propagation optimization method based on niche genetic algorithm according to claim 3 is characterized in that: The congestion propagation influence of each individual in the population is calculated based on the congestion propagation influence of the road section as its fitness value. The specific calculation formula is expressed as: in, It is road section x i In the topology of the Internet of Vehicles, It is road section x i The traffic carrying capacity It is road section x i The number of lanes, It is road section x i , L is the length of the lane occupied by the average motor vehicle for safe driving; The individual congestion propagation influence is obtained as:

5. The multimodal vehicle network congestion propagation optimization method based on niche genetic algorithm according to claim 3 is characterized in that: According to the similarity of the vehicle network road section combination contained in the individual in the vehicle network topology structure, the population is divided into n sub-populations as different microhabitats, including: Apply community detection algorithm to the Internet of Vehicles and divide it into m Leiden communities; For each individual in the population, solve its Internet of Vehicles community belonging vector, the m-dimensional community belonging vector of individual i Among them, the a-th dimension c a represents the total number of road sections that the individual belongs to the connected vehicle community j; Based on the individual Internet of Vehicles community belonging vector, the distance metric between individuals is defined, and the Euclidean distance between individual i and individual j The IoV belonging vector of each individual is used as a cluster. According to the distance metric between individuals, the clusters with the shortest distance are merged into one cluster until n clusters are formed. Finally, the population is divided into n sub-populations as different microhabitats.

6. The multimodal vehicle network congestion propagation optimization method based on niche genetic algorithm according to claim 3 is characterized in that: Randomly select two niches from the divided multiple niches and apply the sorted semi-crossover operator, specifically including: Randomly select two individuals i and j from the population, and sort the road sections included in individual i in ascending order and the road sections included in individual j in descending order according to the influence value of road section congestion propagation; Exchange the first half of the road section of individual i and individual j to form two new individuals i′, j′; At this time, if there are repeated sections in the new individual i′, j′, the repeated sections are removed, and sections are randomly selected from the vehicle network section set to fill in the individual i′, j′ to ensure that the individual i′, j′ still contains k non-repeated sections.

7. A multi-modal vehicle network congestion propagation optimization system based on niche genetic algorithm, characterized in that: A method for optimizing multimodal vehicle network congestion propagation based on a niche genetic algorithm according to any one of claims 1 to 6, the system comprising: a vehicle network information acquisition module, a maximization problem model construction module, and a congestion optimization module; The Internet of Vehicles information acquisition module is used to acquire Internet of Vehicles information; The maximization problem model building module is used to mathematically model the problem of maximizing the influence of vehicle network congestion propagation based on vehicle network information, and obtain the congestion propagation dynamic mechanism and the objective function of maximizing the number of vehicle network congested sections respectively; The congestion optimization module is used to adopt a genetic algorithm based on a fitness sharing niche technology, with maximizing the number of congested sections in the vehicle network as the optimization goal, and calculate multiple heterogeneous congested section combinations that can maximize the number of congested sections in the vehicle network.

8. A computer-readable storage medium storing a program, characterized in that: When the program is executed by the processor, the multimodal vehicle network congestion propagation optimization method based on a niche genetic algorithm as described in any one of claims 1 to 6 is implemented.

9. A computer device comprising a processor and a memory for storing a program executable by the processor, characterized in that: When the processor executes the program stored in the memory, it implements the multimodal vehicle network congestion propagation optimization method based on a niche genetic algorithm as described in any one of claims 1 to 6.

Images